A comparison of features for synthetic speech detection

Sahidullah, Md; Kinnunen, Tomi; Hanilçi, Cemal

doi:10.21437/interspeech.2015-472

Cited by 205 publications

(106 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While we obtained similar results using alternative representations such as linear frequency cepstral coefficients (LFCCs) [7,39], all work reported in this paper was performed with 60dimensional log linear filter bank (LFB) features extracted from 30 ms windows with a 10 ms frame shift and from audio waveforms which are truncated or concatenated to ≈4 second segments (64600 samples). To improve generalisation, we applied frequency masking augmentation [17,40] to mask a random selection of contiguous frequency bands during training.…”

Section: Implementation Detailsmentioning

confidence: 61%

Graph Attention Networks for Anti-Spoofing

Tak¹,

Jung²,

Patino³

et al. 2021

Interspeech 2021

View full text Add to dashboard Cite

The cues needed to detect spoofing attacks against automatic speaker verification are often located in specific spectral subbands or temporal segments. Previous works show the potential to learn these using either spectral or temporal selfattention mechanisms but not the relationships between neighbouring sub-bands or segments. This paper reports our use of graph attention networks (GATs) to model these relationships and to improve spoofing detection performance. GATs leverage a self-attention mechanism over graph structured data to model the data manifold and the relationships between nodes. Our graph is constructed from representations produced by a ResNet. Nodes in the graph represent information either in specific sub-bands or temporal segments. Experiments performed on the ASVspoof 2019 logical access database show that our GAT-based model with temporal attention outperforms all of our baseline single systems. Furthermore, GAT-based systems are complementary to a set of existing systems. The fusion of GAT-based models with more conventional countermeasures delivers a 47% relative improvement in performance compared to the best performing single GAT system.

show abstract

Section: Implementation Detailsmentioning

confidence: 61%

Graph Attention Networks for Anti-Spoofing

Tak¹,

Jung²,

Patino³

et al. 2021

Interspeech 2021

View full text Add to dashboard Cite

show abstract

“…We extract 60-dimensional linear frequency cepstral coefficients (LFCC) [28] of speech as low-level input feature to the neural network. The extraction way is the same as the ASVspoof2019 offical baseline system [27].…”

Section: Methodsmentioning

confidence: 99%

Continual Learning for Fake Audio Detection

Ma,

Yi,

Tao

et al. 2021

Preprint

View full text Add to dashboard Cite

Fake audio attack becomes a major threat to the speaker verification system. Although current detection approaches have achieved promising results on dataset-specific scenarios, they encounter difficulties on unseen spoofing data. Fine-tuning and retraining from scratch have been applied to incorporate new data. However, fine-tuning leads to performance degradation on previous data. Retraining takes a lot of time and computation resources. Besides, previous data are unavailable due to privacy in some situations. To solve the above problems, this paper proposes detecting fake without forgetting, a continual-learningbased method, to make the model learn new spoofing attacks incrementally. A knowledge distillation loss is introduced to loss function to preserve the memory of original model. Supposing the distribution of genuine voice is consistent among different scenarios, an extra embedding similarity loss is used as another constraint to further do a positive sample alignment. Experiments are conducted on the ASVspoof2019 dataset. The results show that our proposed method outperforms fine-tuning by the relative reduction of average equal error rate up to 81.62%.

show abstract

“…Baseline B02 uses linear frequency cepstral coefficients (LFCCs) [18] and a bandwidth of 30 Hz to 8 kHz. LFCCs are extracted using a 512-point discrete Fourier transform applied to windows of 20 ms with 50% overlap.…”

Section: Spoofing Countermeasuresmentioning

confidence: 99%

“…Systems are labelled (left) by the anonymised team identifier (TID) [3]. A short description of each follows: T45 [25]: A light CNN (LCNN) which operates upon LFCC features extracted from the first 600 frames and with the same frontend configuration as the B2 baseline CM [18].…”

Section: Single Systemsmentioning

confidence: 99%

ASVspoof 2019: spoofing countermeasures for the detection of synthesized, converted and replayed speech

Nautsch,

Wang,

Evans

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

The ASVspoof initiative was conceived to spearhead research in anti-spoofing for automatic speaker verification (ASV). This paper describes the third in a series of bi-annual challenges: ASVspoof 2019. With the challenge database and protocols being described elsewhere, the focus of this paper is on results and the top performing single and ensemble system submissions from 62 teams, all of which out-perform the two baseline systems, often by a substantial margin. Deeper analyses shows that performance is dominated by specific conditions involving either specific spoofing attacks or specific acoustic environments. While fusion is shown to be particularly effective for the logical access scenario involving speech synthesis and voice conversion attacks, participants largely struggled to apply fusion successfully for the physical access scenario involving simulated replay attacks. This is likely the result of a lack of system complementarity, while oracle fusion experiments show clear potential to improve performance. Furthermore, while results for simulated data are promising, experiments with real replay data show a substantial gap, most likely due to the presence of additive noise in the latter. This finding, among others, leads to a number of ideas for further research and directions for future editions of the ASVspoof challenge.

show abstract

A comparison of features for synthetic speech detection

Cited by 205 publications

References 36 publications

Graph Attention Networks for Anti-Spoofing

Graph Attention Networks for Anti-Spoofing

Continual Learning for Fake Audio Detection

ASVspoof 2019: spoofing countermeasures for the detection of synthesized, converted and replayed speech

Contact Info

Product

Resources

About